FOLLOWING THE CROWD, ANY EVIDENCE OF HERDING BEHAVIOUR IN THE CRYPTOCURRENCY MARKET

Back to Page Authors: Kofi Agyarko Ababio, Beatrice Desiree Simo-Kengne, N/A

Keywords: Herd behaviour, Bayesian analysis, cryptocurrency, skewness, asymmetry

Abstract: This study investigates the existence of collective action or imitation known as herding behaviour in the cryptocurrency market. Daily data for twenty-one cryptocurrencies are used spanning the period from August 2016 to March 2018. Three market conditions are considered namely bear, normal and bull markets to account for asymmetric effects. Results from Bayesian regressions indicate no herding behaviour in normal conditions. Using 10% as the extreme threshold of market returns, Bayesian regression with asymmetric distributed innovations show evidence of herding under extreme market conditions, particularly during bear and bull markets. Important for portfolio optimisation and management, these findings suggest that symmetric-based approaches are inappropriate for herding detection.